General comments

The manuscript presents one month of hourly measurements of major and trace elements in PM2.5 and PM10 in Warsaw, Poland, in February and March 2016. The data are thoroughly discussed, and three different receptor models are applied to determine the sources and origins of the elements. Five sources in PM10 and seven sources in PM2.5 are found, demonstrating the advantages of high time resolution for appropriate source identification. Furthermore two cases of Saharan dust transport are discussed..

The structure of the manuscript, the results and the presentation of the material are very detailed and carefully worked out. The topic is relevant and well worth publication in ACP. I would, however, suggest a few minor changes and additions before publication.

Specific comments

The advantage of +/- hourly time resolution of elemental concentrations is nicely demonstrated, but biased towards Streaker sampling and PIXE analysis. Recent studies with XRF method have also achieved hourly resolution and size segregation for source apportionment, e.g. see references for Beijing and Delhi in Rai et al. (2021):

Rai, P., Slowik, J. G., Furger, M., El Haddad, I., Visser, S., Tong, Y., Singh, A., Wehrle, G., Kumar, V., Tobler, A. K., Bhattu, D., Wang, L., Ganguly, D., Rastogi, N., Huang, R. J., Necki, J., Cao, J., Tripathi, S. N., Baltensperger, U., and Prévôt, A. S. H.: Highly time-resolved measurements of element concentrations in PM10 and PM2.5: comparison of Delhi, Beijing, London, and Krakow, Atmos. Chem. Phys., 21, 717-730, 10.5194/acp-21-717-2021, 2021. 

In Figs. 4 and 5 the right hand axes are labelled as ‘contribution [%]’, while in the captions you call this ‘explained variation’. Would it not be more consistent (and more correct) to just use ‘contribution’ in both places?

In Fig. 6 the traffic source area is quite different from the road salt source. Shouldn’t we expect more similarity between the two factors, as in both cases probably resuspension would be the mechanism for ‘creating’ the sources? Or is deicing salt not evenly distributed along the Warsaw road system? Please discuss this a bit more.

Fig. 8 shows trajectories originating in or crossing over parts of the Sahara desert and ending above Warsaw in 1500 and 3000 m asl. It is not straight forward that PM arriving at these elevations is measured with ground-based samplers, and possible downward mixing processes should be discussed in more detail. This is especially the case on 18 Feb, where at 500 m asl the airmass is advected from SE, indicating a completely different source location than the Sahara. While I find these two cases interesting and plausible, a strong connection is not given. I recommend adding one or two sentences discussing the uncertainties (Saharan dust composition, vertical mixing from upper layers, concentrations aloft).

In Fig. 9, I do not see an advantage of comparing the elemental time series in ng/m3 with the PMF time series in arbitrary units, as the main elements comprising the soil dust are those in eq. (8), so we basically just add O (in stoichiometric ratios) to these elements. It should be possible from a linear regression to estimate the fraction of the two (I guess it would be something around the value of 2). With respect to the Saharan dust, it would be the same to just compare the five relevant elements, not their oxides, as long as we do not have quantitative information on the amount of dust (mass) transported.

Technical corrections

L118      device (delete s)

L121      substrata (insert s)

L126      beam (delete s)

L131      write Micromatter in one word

L141      Reorder the sentence. Write ‘… is a widely used (…) multivariate factor analysis model in air quality studies based on …’

L142      weighted least squares fit (add an s)

L369      Can you explain the term ‘bioavailable’ a bit better? The way you use it, it appears to be something like a quantitative entity with a time dependence.

L373      … emitting a substantial amount … (insert a)

L377      … attributed to wood combustion … (delete the)

L387      In the case of the coarse …  (insert the)

L393      … the pattern of the source … (insert the)

L504      … points towards two …   (insert towards)

L516      … all ranges of wind … (insert s)

L518      … are used for the maintenance … (replace to with ‘for the’)

L533      write ‘…of the sources identified by PMF for….’

L592      widely (instead of wide)

L633      … receptor modeling based …  (insert ing)

The manuscript presents a study to identify the sources of fine and coarse particulate matter based on high resolution sampling of elemental components in the city of Warsaw. The sample collection and the data treatment were carried out according to state-of-the art methodologies. The manuscript is well organized and complete. The figures are of good quality and summarize the data properly. The discussion of some of the sources needs to be improved as pointed out in the specific comments. The text should be revised by a native speaker.

Specific comments

Line 81: the year should be between parentheses

Line 141: Maybe Belis et al., 2020 (https://doi.org/10.1016/j.aeaoa.2019.100053) is a more appropriate citation for this paragraph

Line 171: The authors should discuss what's the impact of the absence of such important components on the ability of the analysis to quantify sources

Line 222: Please, discuss if the soluble ions are comparable with the elemental determinations

Lines 220 – 228: This paragraph should go to materials and methods

Lines 249-251: This paragraph should go to materials and methods

Line 369: The bioavailability issue should be explained/discussed more in detail

Line 396: Although sulfate is a common component of biomass burning profiles, claiming this fuel is a major contributor to secondary sulfate should be better supported by evidence. Especially in an area where coal combustion, a well-known SO2 source, is well documented.

Line 494: Consider calling this source mixed wood and coal combustion

Line 568: The aged sea salt contribution may be higher than the authors’ estimations considering that what is identified as "road salt" has the highest shares in the three clusters originated in the North Atlantic

Line 580: I guess you mean “m above ground level” please, specify

Lines 615-617: This sentence is not clear, please, rephrase

Line 622: excluding the secondary aerosol

Table 1: you should point out when min values are below the detection limit here

Figures 4 and 5: Please, put the labels of the x axis categories also at the top for a better visualization.

Figure 7: I suggest to use the same colour for corresponding sources in the fine and coarse fraction.

Authors present results of high time resolution measurement of elements in fine and coarse urban aerosols with subsequent identification of sources using a combination of three receptor models.

The paper provides interesting results with detailed insights into the winter pollution sources in Warsaw area.

The paper is suitable for publication in the journal Atmospheric Chemistry and Physics, however, there are a few issues that need to be addressed before acceptance for publication. Minor revisions of the paper taking into consideration the comments reported below are requested.

Comments:

line 19:    Exhaust traffic emissions are mostly prevailing in fine (and especially in submicrometre) fraction.

line 34-35: Statement that the greatest health risk is from PM25 is relative. It depends on considered particle size. The statement is true if you compare PM10 and PM2.5. But it is not true for the comparison of PM2.5 and PM1. PM1 and especially UFP are more dangerous than PM2.5 due to their ability to penetrate deeper into the lung than PM2.5 particles.

line 73:    It is not true, you have overlooked some papers, for example, Pokorná et al., Sci. Total Environ. 2015, 502, 172–183.

line 118: Add sampling flow rate and volume of passed air per sample.

line 281:   Cl originates also from various combustion sources, more details see for example in Mikuška et al, Atmosphere 2020, 11, 688. 

line 363-384: Component contribution and time profile of this factor in coarse fraction suggest considering renaming this factor to residential heating.

line 466-469: Mentioned metals could also indicate emissions from waste incinerator. Is there any incinerator in the vicinity?

line 495-501: Br, Se, As are strong markers of coal combustion, so this factor looks rather like combined combustion of coal and biomass (wood).

line 608-609: As far as I know other studies are providing high time resolution measurement in Central Europe, see comment for line 73.

line 626-627: According to my experience, parallel measurement of different PM fractions with shorter time resolution can also provide the same results as high time resolution measurement.